US 7724817 B2 Abstract A step-size estimator for controlling the step-size of an adaptive equalizer incorporated in a transceiver (e.g., a wireless transmit/receive unit (WTRU)). The step-size estimator updates at least one adaptive equalizer tap used by the adaptive equalizer based on an apparent speed of a channel established between the transceiver and another transceiver. The step-size estimator includes a speed estimator, a signal-to-noise ratio (SNR) averager and a step-size mapping unit. The speed estimator is used to estimate the apparent speed of the channel (i.e., the observed and/or measured rate of change of the channel impulse response). The SNR averager generates a common pilot channel (CPICH) SNR estimate. The step-size mapping unit uses the speed estimate and the CPICH SNR estimate to generate a step-size parameter, μ, and a filter taps leakage factor parameter, α, used by the adaptive equalizer to update the filter tap coefficient.
Claims(11) 1. A step-size estimator for controlling the step-size of an adaptive equalizer comprising:
a delayed locked loop (DLL) for generating a punctual sample sequence based on received samples and first significant tap (FSP) location information, wherein the DLL comprises:
a delay buffer for generating delayed samples based on the received samples;
an interpolator for shifting the delayed samples, generating early and late time sequences, and generating the punctual sample sequence;
a code tracking loop (CTL) for generating an interpolation index signal and a buffer address signal;
an adder for generating an index signal by adding together a given FSP location signal and the buffer address signal, wherein the delay buffer aligns received samples for tracked FSPs within a certain resolution based on the index signal; and
a common pilot channel (CPICH) signal-to-noise ratio (SNR) estimator for generating a CPICH SNR estimate based on the punctual sample sequence, wherein at least one filter tap coefficient used by the adaptive equalizer is updated based on an apparent speed of a channel.
2. The step-size estimator of
a step-size mapping unit for mapping the CPICH SNR estimate to at least one parameter used to update at least one filter tap coefficient associated with the adaptive equalizer.
3. The step-size estimator of
speed estimator for estimating the apparent speed of the channel based on the punctual sample sequence, wherein the step-size mapping unit uses the apparent speed estimate and the CPICH SNR estimate to generate the at least one parameter.
4. The step-size estimator of
5. The step-size estimator of
6. A method of controlling the step-size of an adaptive equalizer comprising:
generating a punctual sample sequence based on received samples and first significant tap (FSP) location information, wherein the generating comprises:
generating delayed samples based on the received samples;
shifting the delayed samples, generating early and late time sequences, and generating the punctual sample sequence;
generating an interpolation index signal and a buffer address signal;
generating an index signal by adding together a given FSP location signal and the buffer address signal, wherein the received samples are aligned for tracked FSPs within a certain resolution based on the index signal; and
generating a common pilot channel (CPICH) signal-to-noise ratio (SNR) estimate based on the punctual sample sequence, wherein at least one filter tap coefficient used by the adaptive equalizer is updated based on an apparent speed of a channel.
7. The method of
mapping the CPICH SNR estimate to at least one parameter used to update at least one filter tap coefficient associated with the adaptive equalizer.
8. The method of
estimating the apparent speed of the channel based on the punctual sample sequence; and
using the apparent speed estimate and the CPICH SNR estimate to generate the at least one parameter.
9. The method of
10. The method of
11. An integrated circuit (IC) for controlling the step-size of an adaptive equalizer comprising:
a delayed locked loop (DLL) for generating a punctual sample sequence based on received samples and first significant tap (FSP) location information, wherein the DLL comprises:
a delay buffer for generating delayed samples based on the received samples;
an interpolator for shifting the delayed samples, generating early and late time sequences, and generating the punctual sample sequence;
a code tracking loon (CTL) for generating an interpolation index signal and a buffer address signal;
an adder for generating an index signal by adding together a given FSP location signal and the buffer address signal, wherein the delay buffer aligns received samples for tracked FSPs within a certain resolution based on the index signal; and
a common pilot channel (CPICH) signal-to-noise ratio (SNR) estimator for generating a CPICH SNR estimate based on the punctual sample sequence, wherein at least one filter tap coefficient used by the adaptive equalizer is updated based on an apparent speed of a channel.
Description This application is a divisional of U.S. patent application Ser. No. 11/238,469, filed Sep. 29, 2005, which claims the benefit of U.S. Provisional Application No. 60/625,869, filed Nov. 8, 2004, which is incorporated by reference as if fully set forth. The present invention relates to controlling an adaptive equalizer incorporated in a transceiver, such as a wireless transmit/receive unit (WTRU). More particularly, the present invention relates to updating at least one filter tap coefficient used by the adaptive filter based on the apparent speed of a channel (i.e., the observed and/or measured rate of change of the channel impulse response) established between the transceiver and another transceiver. An adaptive equalizer based receiver, such as a normalized least mean square (NLMS)-based receiver, provides superior performance for high data rate services such as frequency division duplex (FDD) high speed downlink packet access (HSDPA) or code division multiple access (CDMA) 2000 evolution data voice (EV-DV) over a Rake receiver. A typical NLMS receiver includes an adaptive equalizer having an equalizer filter and a tap coefficients generator to generate the tap coefficients used to update the filter coefficients of the equalizer filter. The equalizer filter is typically a finite impulse response (FIR) filter. An adaptive step-size parameter, μ, (“mu”) in an adaptive equalization algorithm controls the rate of convergence of the equalizer filter. The adaptation step-size parameter μ is a critical parameter that impacts the performance of the adaptive equalizer. The adaptive step-size parameter μ is typically defined prior to operation of the equalizer filter or varied in a deterministic way. The step-size is the size of each step in an iterative (loop) algorithm that attempts to converge to some point, such as least mean square (LMS), NLMS or its derivatives. Large step-sizes help the adaptive equalizer converge (in as accurate a manner as is possible) in a short period of time, but the adaptive equalizer would converge more accurately if the step-size was smaller. Thus, there is a trade-off between quick and accurate convergence. The ideal balance between convergence speed and accuracy depends on how fast the point on which the algorithm is trying to converge to is changing. The convergence time is inversely related to the adaptation step-size parameter μ. Therefore, with a larger step-size, the convergence may be obtained quickly. However, the large step-size may cause misadjustment errors which impact the raw bit error rate (BER) performance of the adaptive equalizer. The misadjustment errors are due to the convergence of the LMS never being fully achieved because the step size used is approximately the closest each point on the vector may come to the desired point. The present invention is a step-size estimator for controlling the step-size of an adaptive equalizer incorporated in a transceiver (e.g., a WTRU). The step-size estimator updates at least one adaptive equalizer tap used by the adaptive equalizer based on an apparent speed of a channel established between the transceiver and another transceiver. The step-size estimator includes a speed estimator, a signal-to-noise ratio (SNR) averager, and a step-size mapping unit. The speed estimator is used to estimate the apparent speed of the channel (i.e., the observed and/or measured rate of change of the channel impulse response). The SNR averager generates a common pilot channel (CPICH) SNR estimate. The step-size mapping unit uses the speed estimate and the CPICH SNR estimate to generate a step-size parameter, μ, and a filter taps leakage factor parameter, α, used by the adaptive equalizer to update the filter tap coefficient. A more detailed understanding of the invention may be had from the following description, given by way of example and to be understood in conjunction with the accompanying drawings wherein: When referred to hereafter, the terminology “WTRU” includes but is not limited to a user equipment (UE), a mobile station, a fixed or mobile subscriber unit, a pager, or any other type of device capable of operating in a wireless environment. When referred to hereafter, the terminology “transceiver” includes, but is not limited to, a base station, a WTRU, a Node-B, an access point (AP) or any other wireless communication device that receives signals from and transmits signals to another transceiver. When referred to hereafter, the terminology “apparent channel speed” and “apparent speed of a channel” includes, but is not limited to, the observed and/or measured rate of change of an impulse response of a channel established between a first transceiver (e.g., WTRU, base station, or the like), and at least one other transceiver. The change of the channel impulse response may be caused by the movement of one or more of the transceivers, oscillator error which occur in at least one of the transceivers, and the movement of objects in the environment in which at least one of the transceivers operates. The features of the present invention may be incorporated into an integrated circuit (IC) or be configured in a circuit comprising a multitude of interconnecting components. The present invention controls the adaptation step-size of an adaptive equalizer. The value of the adaptation step-size μ depends on the rate of channel change (such as a Doppler spread which is related to the WTRU velocity), and SNR of the channel. For fast channels, it is preferable to use a larger step-size to allow the adaptive equalizer to track the channel variations quickly. Conversely, for slower channels, a lower step-size is desired to reduce the misadjustment error, and thus improve the performance of the adaptive equalizer. The dependency of the adaptation step-size parameter μ on the SNR is such that at a high SNR, the value of the adaptation step-size parameter μ tends to be higher, while at a low SNR, the adaptive step-size parameter μ is typically small. Additional inputs may also be used as appropriate (e.g., delay spread and the number of active taps in the equalizer filter). The present invention is used to maintain an ideal balance between the convergence speed and accuracy through the estimation of the apparent channel speed. Referring to When the second transceiver As shown in Referring to In accordance with the present invention, velocity information is extracted from a history of the filter coefficients used by the equalizer tap update unit The tap coefficient extractor A typical channel impulse response can usually be characterized by finite set of (disjoint) delayed and scaled impulses. The location of each of these impulses is referred to as a path (i.e., a component of a “multi-path” channel). The location and the mean power of each of the paths relative to a first significant tap (FSP) determine the location and magnitude of the equalizer tap weights. The extracted tap coefficient The full length of the TDL A vector of delays The phase difference function generator The averaging filter The normalizing unit For example, the normalization is performed by dividing each element of the average phase difference vector Each element of the normalized phase difference function vector The goal of the delay calculator The threshold delay The mapping from speed and SNR is determined empirically. This is done by simulating the performance of the receiver with various values of the step-size, μ (“mu”), parameter The filter taps leakage factor, α, is defined as follows:
The adaptation of the filter coefficients in a generic LMS algorithm can be written as:
Referring to The CPICH SNR estimator The estimated CPICH SNR The received samples The delay buffer The early and late sample sequences The interpolator The interpolator The size of the delay buffer In accordance with this embodiment, the amount of delay required between a current symbol and delayed symbols to achieve a target phase in the delay buffer The punctual (i.e., on-time) sample sequence The symbol sequence The complex conjugate The control loop The normalization process is a necessity in any case to ensure repeatability for the speed in different signal to noise ratios. The filtered conjugate signal The resulting normalization values range from 0 to 1. The minimum delay Since the correlations are performed by using a known sequence (i.e., CPICH signal), the SNR level of the correlated signal will have direct impact on the calculated correlations. A reference/correlation value signal The normalization forces the quotient result signal The loop filter Therefore, the speed is inversely proportional to the delay amount to set the normalized auto-correlation to the reference level The control loop The present invention is based on the fact that the autocorrelation function for the Doppler spectrum is a 0th order Bessel function. The Bessel behavior permits a correlation value to be set to estimate the amount of delay to achieve desired correlation between a current symbol and delayed symbol. As shown in The optional mapping units While the present invention has been described in terms of the preferred embodiment, other variations which are within the scope of the invention as outlined in the claims below will be apparent to those skilled in the art. Patent Citations
Referenced by
Classifications
Legal Events
Rotate |